Colon cancer is a heterogeneous disease whose clinical behavior is incompletely predicted by the present staging system. A precise understanding of the biology of individual tumors is required for both appropriate clinical decision-making and development of new cancer therapies. The goal of our work is to characterize an individual patient's tumor in sufficient detail to allow selection of optimal therapy. By studying tumor markers that illustrate differences in the function of the tumor cells compared to normal tissue, we will also identify biologic processes that are potential targets for new cancer treatments. Achieving this goal involves two main activities. First, subgroups of patients within the present staging system that respond either well or poorly to current adjuvant therapies must be identified. There are a large number of tumor-specific prognostic factors available to meet this goal, but because these have never been studied in combination in large prospective clinical trials, their utility is not proven. Second, we must further our study of colon cancer biology through preserving a resource that will allow evaluation of new markers as they are developed. Unless many different markers are studied in relation to each other, the importance of genes associated with genomic instability, cell cycle progression, tumor metastasis and resistance to therapy in patients participating in Intergroup adjuvant therapy trials for either Stage II or Stage III colon cancer (CALGB 9581 and 89803, respectively) and to correlate these values with time to recurrence and overall survival. We will also develop and validate a new method of measuring gene expression in fixed, embedded tissue which will markedly expand the ability to correlate marker expression and outcome in cooperative group cancer trial patients. The results of these studies will have great clinical importance as it will allow clinicians to avoid toxic and costly treatment of patients unlikely to benefit from this and will help to tailor adjuvant therapy to individual patients based upon a molecular marker profile and potentially, and may reveal targets for the development of new therapies.